College Park, Maryland June 6 - 10 , 2004 |
M2-C5 (11:45 AM): Instrument Design via Envelope Back and Spreadsheet
J.F. Ankner, Ch. Rehm (Spallation Neutron Source, Oak Ridge National Laboratory)
All too often, the first recourse of the neutron instrument designer is to some sort of stochastic or Monte Carlo (MC) simulation. While very efficient at generating results, MC simulation can suffer from a deficiency of insight, failing to provide clear criteria to direct instrument design choices. This state of affairs is as unnecessary as it is unfortunate, since the modules used in MC simulation must, by definition, contain analytical descriptions of the components simulated. By abstracting out the analytical formalism, thinking a bit about the relevant phase space, and developing common-sense metrics, one can provide a framework for understanding simulation results. Further, since these computations generally consume far less time than full-blown MC, one can model phenomena, such as background scattering, that simply cannot practically be addressed via MC. We will describe our application of acceptance-diagram techniques to primary neutron optical design (including disk choppers) and present a set of phase-space simulation metrics for evaluating results. In addition, we will consider the optical characteristics of neutron scattering/emission from interior beamline surfaces, a type of secondary-source problem difficult to treat using MC. Ultimately, the surest designs arise from judicious use of both analytical and stochastic methods, the former providing a framework for understanding the latter, as well as the means for circumventing their computational limitations.
Back to the Program